There is a growing demand for affordable mid-infrared sources for use in a variety of applications including atmospheric sensing (global wind sensing and low altitude wind shear detection), eye-safe medical laser sources for non-invasive medical diagnostics, eye-safe laser radar and remote sensing of atmospheric constituents, optical communication, and numerous military applications such as target designation, obstacle avoidance and infrared counter measures. These applications rely on the existence of “spectroscopic fingerprints” of numerous organic molecules in the mid-IR range.
Recent research advances have spurred considerable effort in the development of practical mid-IR sources. This work has included direct generation in semiconductors using InAsSbP/InAsSb/InAs,1 and quantum cascade lasers2. Mid-IR wavelengths have also been generated using nonlinearities in Optical Parametric Oscillators3 and difference frequency generators.4,5 All of these approaches yield tunable sources in the mid-IR and all suffer some fundamental problems that limit their use as robust low cost mid-IR source. Furthermore, to date, all of these sources have limited output powers that preclude their use in higher power applications such as remote sensing.
In contrast to the relatively large body of work using the approaches described above, there has been relatively little investigation of the potential for direct oscillation from divalent transitional metal ions (TM2+) placed in the asymmetric (Td) lattice sites of the wide bandgap binary and mixed ternary II-VI semiconductor crystals. The lack of work on direct emission of chromium doped (or other transitional metal doped) sources in the mid-IR has one primary cause. Long wavelength TM emissions are quenched by multi-photon processes in conventional laser host media such as oxide and fluoride crystals, resulting in extremely low room-temperature quantum efficiency of fluorescence.
Recently, mid-IR laser activity near 2-4 μm was reported for Cr:ZnS6,7,8,9,10, Cr:ZnSe6,7,11,12,13,14,15,16,17,18, Cr:Cd1-xMnxTe19, Cr:CdSe20, and Fe2+:ZnSe21 crystals. These TM doped II-VI compounds have a wide bandgap and possess several important features that distinguish them from other oxide and fluoride laser crystals. First is the existence of chemically stable divalent TM dopant ions, which substitute Zn2+ or Cd2+ host ions, with no need for charge compensation. An additional feature of the II-VI compounds is their tendency to crystallize in tetrahedrally coordinated structures. As opposed to the typical octahedral coordination at the dopant site, tetrahedral coordination gives smaller crystal field splitting, placing the dopant transitions further into the IR. Finally, a key feature of these materials is a poor phonon spectrum that makes them transparent in a wide spectral region, decreases the efficiency of non-radiative decay and gives promise to a high yield of fluorescence at room temperature.
In terms of merit for high average power applications, it is known that some of chalcogenides (e.g. ZnS and ZnSe) feature excellent thermo-mechanical properties, having thermal shock resistance values comparable to and coefficient of thermal conductivity better than such thermo-mechanically robust materials as YAG crystals. Given the attractive thermo-mechanical, spectroscopic properties of TM2+, and nice overlap of the Cr2+ absorption and emission Er and Tm fiber lasers as well as of stained layer InGaAsP/InP and, theoretically, InGaNAs/GaAs diode lasers, directly fiber or diode-pumped wide band semiconductor crystals doped with TM ions can be considered as very promising and effective systems for medicine, remote sensing, trace gas analysis, and high power wavelength specific military applications.
The studies of TM2+ doped II-VI materials showed that in terms of spectroscopic and laser characteristics these media are very close mid-IR analogues of the titanium-doped sapphire (Ti—S). It is anticipated that, similarly to the Ti—S laser, TM2+ doped chalcogenides will be lasing in the near future with a great variety of possible regimes of oscillations, but with an additional significant advantage of being directly pumpable with radiation of InGaAsP or InGaNAs diode arrays.
During the last 2-7 years several groups, including the inventors, have actively explored analogues TM2+ crystal hosts for tunable lasing in CW, free-running long pulse, Q-switched and mode-locked regimes of operation. So far the most impressive results—room temperature operation, >60% lasing efficiency, 3.7 W of output power, more than 1000 nm range of tunability—have been obtained using Cr2+:ZnSe crystals. Based on these results, it appears that Cr doped ZnS and ZnSe crystals possess a unique combination of technological, thermo- mechanical, spectroscopic, and laser characteristics that make them potentially low cost, affordable mid-IR laser sources.
However, in these spectroscopic and laser studies of TM2+:II-VI materials there was no indication that microchip lasers and chip-scaled integrated lasers could be designed on the basis of TM doped II-VI hosts. Microchip lasing requires several specific factors in addition to standard factors required for any laser media. These additional factors are high optical density and high gain of thin layers (usually <1-2 mm) of active material, which is translated into high cross sections of absorption and emission, combined with a high doping levels of active ions at which there is still no concentration quenching of fluorescence and no degradation of the optical quality of the host material.
Also unknown in the prior art is a design of “spatially dispersive” cavities for realization of flexible laser modules easily reprogrammable from monochromatic to ultrabroadband and multiline regimes of operation.
U.S. Pat. Nos. 5,461,635 and 6,236,666 taught the approach of superbroadband (SBL) or multiwavelength system22,23,24,25 based on spatial separation of different wavelengths in a single laser cavity. The optical components of the cavity maintain distinct gain channels in the active zone of semiconductor chip, reduce cross talk, suppress mode competition, and force each channel to lase at a specific stabilized wavelength. By designing this cavity structure appropriately, the system creates its own microcavities each lasing at different wavelengths across the complete gain spectrum of the active material. The system is ideal from the point of view of control of laser wavelengths generated in a common laser cavity and allows the obtaining of very small and controllable wavelength spacing. This approach allows the construction of a laser that emits a plurality of narrow spectral lines that can be easily tailored to any pre-assigned spectral composition within the amplification spectrum of the gain medium. This approach has been demonstrated for the emission of thirty lines in laboratory conditions and the stability and line width measurements are extremely promising. Conventional tunable laser systems used for remote sensing are appropriate only for single element analysis. Proposed simple, flexible and easily reprogrammable laser modules open new opportunities for simultaneous multi-element gas tracing analysis. It appears that TM doped II-VI hosts and, specifically, chromium doped ZnS and ZnSe crystals featuring broad amplification spectra are ideal active media for superbroadband and multiline lasing.
Finally, the prior art has not taught utilization of acousto-optic, electro-optic, photorefractive and birefringent properties of II-VI crystals in one integrated microchip system combining active medium, acousto- or electro-optic modulator, filter, other passive components of the cavity such as waveguide grating, or birefringent filter.